Supplementary Info

[ngreenwald]

Supplementary Figures:

• [x] Panel validation @alex-l-kong
  • [x] Representative FOV(s) with all channels concurrently
  • [x] One control tissue, one tumor tissue?
• [x] ROI selection @srivarra
  • [x] Show H&E stain of a given core, matching MIBI image. Map MIBI image back to H&E with boxed diagram. Pick a couple representative FOVs.
  • [x] Number of ROIs per timepoint
• [x] QC @camisowers
  • [x] Longitudinal controls: show consistency of staining across longitudinal samples over all TMAs, along with example images
  • [x] TMA location: Show that there are no significant batch effects
• [x] Image processing
  • [x] Rosetta @cliu72
  • [x] Cartoon diagram showing the workflow
  • [x] Iterative selection of coefficients
  • [x] Example images of input channels, target channel, and output @alex-l-kong
  • [x] Normalization @cliu72
  • [x] Calculating MPH
  • [x] Constructing a normalization curve
  • [x] Fitting a per-run per-mass coefficient
  • [x] Normalizing using that coefficient
  • [x] Example tiled image in acquisition order before and after @alex-l-kong
• [x] Cell identification and classification
  • [x] Segmentation: single-channel TIFs of each segmentation channel, the combined channel, the output, the overlay, for multiple FOVs @srivarra
  • [x] Clustering @camisowers
  • Number of cells per cell cluster
  • Number of cells per FOV
  • CPMs from random subset of 20 FOVs?
  • Cell type composition by tissue location of met
  • Flowchart showing all three levels of cell clustering
• [x] Functional marker thresholding: @alex-l-kong
  • [x] histogram of marker expression for all functional markers, with line indicating threshold for positivity
  • [x] Example screenshot of Mantis showing how thresholds are identified
• [x] Feature extraction @camisowers
  • Defining image compartments: inputs, smoothing, output mask
  • Computing density: counts per image, background mask, compute area
  • Ratios: list of major cell types with all pairwise combinations
  • Minimum criteria: # of cells, # of positive cells, etc for feature to be computed
  • Mixing score
  • ECM features
  • Correlation filtering

Supplementary Tables

• [ ] Antibody information
• [ ] Ranked list of top hits
• [ ] Clinical information

[ngreenwald]

We can discuss more details about each task tomorrow, as well as do any swaps you guys want.

[ngreenwald]

Plotting code can go here: https://github.com/angelolab/TNBC_python_scripts/blob/main/python_files/generate_supplementary_plots.py

[ngreenwald]

Boom!